17549768. VECTORIZED BILINEAR SHIFT FOR REPLACING GRID SAMPLING IN OPTICAL FLOW ESTIMATION simplified abstract (QUALCOMM Incorporated)

From WikiPatents
Jump to navigation Jump to search

VECTORIZED BILINEAR SHIFT FOR REPLACING GRID SAMPLING IN OPTICAL FLOW ESTIMATION

Organization Name

QUALCOMM Incorporated

Inventor(s)

Rajeswaran Chockalingapuramravindran of San Diego CA (US)

Kristopher Urquhart of Poway CA (US)

Jamie Menjay Lin of San Diego CA (US)

Risheek Garrepalli of San Diego CA (US)

VECTORIZED BILINEAR SHIFT FOR REPLACING GRID SAMPLING IN OPTICAL FLOW ESTIMATION - A simplified explanation of the abstract

This abstract first appeared for US patent application 17549768 titled 'VECTORIZED BILINEAR SHIFT FOR REPLACING GRID SAMPLING IN OPTICAL FLOW ESTIMATION

Simplified Explanation

The abstract of the patent application describes a computer-implemented method for generating a vectorized bilinear shift of an input. Here is a simplified explanation of the abstract:

  • The method receives a first input.
  • The first input is interpolated using a first shift along a first dimension and a second shift along a second dimension.
  • Based on the interpolated first input, a first output is generated.
  • The first output can be used as a replacement for grid sampling algorithms.

Potential Applications of this Technology:

  • Image processing and computer vision: The method can be used to enhance image quality and manipulate images in various applications such as image recognition, object detection, and video processing.
  • Graphics rendering: The method can improve the rendering of graphics in video games, virtual reality, and computer-generated imagery (CGI).
  • Machine learning: The method can be applied in training and inference processes of machine learning models, improving their accuracy and efficiency.

Problems Solved by this Technology:

  • Grid sampling algorithms can be computationally expensive and may introduce artifacts or distortions in the output.
  • Traditional interpolation methods may not provide the desired level of accuracy or smoothness in the shifted input.
  • The method solves these problems by generating a vectorized bilinear shift of the input, which can be more efficient and produce higher-quality results.

Benefits of this Technology:

  • Improved efficiency: The method offers a faster alternative to grid sampling algorithms, reducing computational requirements and processing time.
  • Enhanced output quality: The vectorized bilinear shift provides a more accurate and smoother representation of the input, minimizing artifacts and distortions.
  • Versatility: The method can be applied to various types of inputs, making it suitable for a wide range of applications in image processing, graphics rendering, and machine learning.


Original Abstract Submitted

A computer-implemented method includes receiving a first input. The first input is interpolated based on a first shift along a first dimension and a second shift along a second dimension. A first output is generated based on the interpolated first input. The first output corresponds to a vectorized bilinear shift of the first input for use in place of grid sampling algorithms.